Beta, v0.3.0-beta.8 · Python, TypeScript

AXIOM AI

Check your model, data and computer before you start.

Most AI projects fail early for boring reasons: the model doesn’t fit, the dataset has broken rows, or the setup is wrong. AXIOM catches those problems in seconds, on your own machine, before you spend hours of GPU time.

What it does

Models

Inspect local folders or Hugging Face repos without downloading weights.

Datasets

Validate JSONL, find duplicates, estimate tokens and write a cleaned copy.

Hardware

Detect CPU, RAM, GPU and VRAM, then estimate which model sizes fit.

Training

Hardware-aware LoRA and QLoRA plans, plus optional local fine-tuning.

MCP server

Default-deny tools so AI agents only touch what you allow.

Control Center

A browser interface for everything, started with one command.

Try it

python -m pip install --upgrade --pre axiom-all
axiom init my-ai
cd my-ai
axiom doctor
axiom train plan 7 --method qlora
axiom gui

Local by default. Plans are estimates, and real results depend on your model, data and hardware.

Live repo stats

starsissuesrelease

Built with